estimation of integral time scale

Clementine

Hi everyone, I am currently running unsteady, LES simulations on fluent. But I can not see the effects of the turbulence because my time step is higher than the integral time scale=T. When I go through the fluent help, they estimate the integral time scale (21.2.2) thanks to the rms velocity or to k and epsilon. But each time that I try to estimate T with the rms velocity, its value depends on my time step. So it is wrong. And to estimate T with k, and epsilon, i don't know where we can have access to the values of k and epsilon.

If someone can give me a hint